Codecs for film delivery

Jessariah67 wrote on 1/10/2020, 6:34 PM

Hey All,

I've been around here and on Creative Cow, but I'm not getting where I need to go.

Our delivery schedule is calling for the master file to be"ProRes 422 HQ or HDCAM SR 1080p." I've read that the Magix ProRes-turned-Magix Intermediate is a "fake" ProRes. I rendered a 2-minute clip in HDCAM SR 422 and it was 6G in size (using "Default" color space, if that makes a difference), which would mean my movie would be 1/4T in size...

Has anyone found success down this road, with what's available in VP(17)?

I also have the Adobe subscription, so I have Media Encoder as a possible part of the mix/solution.

Any help with any of this would be appreciated.

Thanks!

KH

Comments

Musicvid wrote on 1/10/2020, 7:34 PM

The word "fake" is misleading, it just doesn't use the Apple encoder. Suggest you render one of each, show them to your producer, and arrive a mutually agreeable decision. Yes, the files are large.

john_dennis wrote on 1/10/2020, 8:00 PM

Is someone in your chain going to prepare a Digital Cinema Package (DCP)?

For your casual reading.

I'm rendering last years video, so two more won't make much difference.

When I put both files back on the timeline and compare each to the original, I can't see any difference.

If they will take HDCAM SR, I'd make a trip to Best Buy and deliver it on a $50 hard drive. Or a pair of $50 hard drives. It's a cost of doing business.

Musicvid's right that what they really, really want is all that matters.

john_dennis wrote on 1/10/2020, 8:08 PM

Please. Could someone fix the server?

john_dennis wrote on 1/11/2020, 9:28 AM

I ran picture quality tests on these two formats using Wayne Waag's Render Quality Metrics Tool and found HDCAM SR 422 is more faithful to the original in an objective test.

You decide if the difference in picture quality is worth the difference in file size.

Musicvid wrote on 1/11/2020, 9:32 AM

I ran picture quality tests on these two formats using Wayne Waag's Render Quality Metrics Tool and found HDCAM SR 422 is more faithful to the original in an objective test.

You decide if the difference in picture quality is worth the difference in file size.

I suspect those slight differences would be magnified significantly if using 10 bit source.

john_dennis wrote on 1/11/2020, 9:44 AM

Since I don't know the slope of the curve, I don't know how slight the differences are. I started a project to approximate the curve, but I don't think I have enough life left to finish it.

For playing locally and distribution to family members, I've been using Magix AVC AMD VCE that produces these numbers.

Date: 2020/01/01  22:45:54  
Description: Magix AMD VCE UHD @ 40Mbps
Frames Processed: 914
Mean Squared Error: 5.039
Peak Signal to Noise Ratio: 47.419

Date: 2020/01/02  13:10:34  
Description: Magix AMD VCE FHD 28 Mbps
Frames Processed: 914
Mean Squared Error: 5.414
Peak Signal to Noise Ratio: 46.934
Date: 2020/01/03  15:15:58 

Musicvid wrote on 1/11/2020, 11:30 AM

Since I don't know the slope of the curve, I don't know how slight the differences are. I started a project to approximate the curve, but I don't think I have enough life left to finish it.

Another name for the slope of a nonlinear curve is the derivative, my precalculus skills having been among the first to fall into disuse following retirement. The curves can be plotted relatively easily in Excel, however, as this foam baby gives testament. That's my Precision bass solo at 3:30.

john_dennis wrote on 1/11/2020, 12:18 PM

My precalculus skills having been among the first to fall into disuse following leaving the class. I interviewed for an engineering job in1978. The interviewer had an integral calculus problem on his white board. I declined the job, which, by the way, would have required that I move to Charlotte, NC. I’m undecided if I was put off more by the relocation or the math, but I don’t remember what was on the whiteboards in any of my other interviews.

Jessariah67 wrote on 1/11/2020, 3:10 PM

This has been a lot of help. Thanks!

When I was researching this, there was talk about Magix providing certain codecs that we need for this kind of delivery, but that most users wouldn't have a use for, so why spend more for the licensing? My argument with that is, if you want to make VP a more viable "pro" option, this is the kind of thing that will make a difference in the long run - in line with the newer features that are moving toward team-oriented work flows, etc.

Just part of the growing pains I guess - lijke back in the SF days, when they added "video" to their Vegas DAW and had a lot of composers barking about paying for something they weren't going to use.

adis-a3097 wrote on 1/12/2020, 10:45 AM

Since I don't know the slope of the curve, I don't know how slight the differences are. I started a project to approximate the curve, but I don't think I have enough life left to finish it.

Another name for the slope of a nonlinear curve is the derivative, my precalculus skills having been among the first to fall into disuse following retirement. The curves can be plotted relatively easily in Excel, however, as this foam baby gives testament. That's my Precision bass solo at 3:30.

Fretless? :)

Musicvid wrote on 1/12/2020, 6:50 PM

Jeez, you had to ask.

Yamaha SG key bass on an old S80. My wheel made it "fretless". I was also covering strings, flute, B3, Rhodes, and fx on the RH split. We also had piano, guitar, and a drummer I kept for 10 years. But remember it was 2004.

BTW, and only because the DVD content was "cartoony," I oversharpened in Lanczos and upscaled to 720p. It's good for limited application.

adis-a3097 wrote on 1/13/2020, 5:29 AM

Jeez, you had to ask.

Yamaha SG key bass on an old S80. My wheel made it "fretless". I was also covering strings, flute, B3, Rhodes, and fx on the RH split. We also had piano, guitar, and a drummer I kept for 10 years. But remember it was 2004.

BTW, and only because the DVD content was "cartoony," I oversharpened in Lanczos and upscaled to 720p. It's good for limited application.

Well, being a bass player myself I guess I had to.😂

And I must say - you got me, I thought I was hearing the real strings there. Hats off, Musicvid!

Musicvid wrote on 1/13/2020, 8:16 AM

For bass, keys, percussion, and reedless winds, the SG waveset can't be beat, still use them. Saxes and brass are toxic, though.

I learned LH independence as a B3 player with a TOP/ AWB cover band in the 70s. Real B3, real Rhodes, real Rhodes key bass, and 1Kw Leslies. Never really mastered the DX7.

vkmast wrote on 1/13/2020, 9:43 AM

@Musicvid I guess this would be too far from your area to attend? (Or too late?) 😎

Musicvid wrote on 1/13/2020, 10:27 AM

Mmm.. Squibb Cakes!

Dave Garibaldi is the greatest funk drummer of all time --had the privilege of working with one of his proteges' (Fred Weber) on tour. Not quite the same without Chester Thompson on B3, though. AWB sounds just like they did forty years ago. Our band followed The Spinners ("Rubber Band Man") on the Seaway Hotel tour in Canada. Hung out with Philip and Andrew in our dorm and on Saturdays at the Catacombs in Boulder, pre-EWF. Ginsberg kept asking me out, but that's another story.

It was all cut short, though, when all the great dance clubs in the Midwest became discos overnight in '78 and we found ourselves stranded in Pontiac without a contract. Wouldn't trade the experience for anything, but would never do it again.